Quantum Computing Is Coming For Your Data, This Startup Thinks It Has A Solution

Up until recently, conversations around artificial intelligence have focused on what AI can do and how it can change our lives. But as businesses rush to adopt the advanced technology, the question that’s becoming increasingly urgent is centred on data privacy. That is, how are organisations using AI safely, without putting sensitive data at risk?

Integrated Quantum Technologies (IQT) is aiming to solve this issue. Described as building quantum-ready infrastructure for enterprises, IQT has developed a platform that’s designed to help organisations deploy AI and unlock the full value of their most sensitive data, without ever exposing it. At the heart of its offering is a product called VEIL, short for Vector Encoded Information Layer.

The problem at hand is one that is undoubtedly complex and requires a solution that is able to rise to the occasion. IQT has managed to build something up to par while ensuring its implementation process is simple and smooth.

AI Has A Data Problem

Modern AI systems thrive on data – the more information they can access, the more useful they become.

The trouble, however, is that a lot of the data that organisations would most like to use is also the data they’re least comfortable sharing. Things like financial records, healthcare information, customer details and proprietary business data all present significant privacy and security concerns.

Historically, businesses have relied on privacy and security techniques like encryption, homomorphic encryption and differential privacy to protect this information. But the issue, according to IQT, is that many of these approaches force organisations to make trade-offs between security, performance and cost. And as AI adoption accelerates, those compromises are becoming a lot more difficult to justify.

Is the Quantum Threat Closer Than We Think?

Part of the urgency stems from the arrival of quantum computing. Although large-scale quantum computers capable of breaking modern encryption may still be years away, security experts are increasingly warning about a strategy known as “Harvest Now, Decrypt Later”.

The concept is straightforward: bad actors intercept and store encrypted data today, with the expectation that future quantum computers will eventually be powerful enough to decrypt it. In other words, data that appears secure right now may not remain secure forever.

Kind of like how police and other authorities used to collect DNA samples from crime scenes long before we had the technology to analyse it. They kept it anyway, just in case we’d be able to figure it out one day. And, low and behold, we did!

So for businesses handling sensitive information, that presents a potentially significant long-term risk, both financially and reputationally.

A Different Approach To Data Protection

Rather than focusing on encrypting data, VEIL takes a different route. The platform transforms data into compressed, anonymised representations that retain the predictive value needed for AI and machine learning systems whilst removing sensitive information. According to IQT, these transformed representations are irreversible, which means that the original raw data actually can’t be reconstructed. From there, AI models work with the transformed data rather than the underlying information itself.

The result of this is that organisations can train and deploy AI systems without exposing the raw data that privacy regulations are designed to protect.

This is a notable departure from traditional approaches, in which data is often protected in one form but can become exposed during processing.

Solving A Global Compliance Challenge

The technology could also help address one of the biggest operational headaches facing multinational organisations today and in the future. That is, many companies currently deploy separate AI systems in different regions to comply with local data residency requirements. This creates duplication, increases costs and can lead to inconsistencies between models operating in different markets.

IQT shares that VEIL allows organisations to operate a single AI instance globally without requiring region-specific duplication. If this is successful, that could reduce operational complexity whilst helping businesses maintain compliance across multiple jurisdictions.

Where To From Here?

Whether VEIL becomes the industry standard remains to be seen, but the company’s approach reflects something a lot bigger and more meaningful on an industry level. The conversation is no longer just about building more powerful AI models; now, it’s about building systems that organisations can actually trust beyond the here and now.

As concerns around privacy, regulation and quantum-era security continue to grow, companies that can remove the trade-off between innovation and protection may find themselves in a particularly strong position. After all, the biggest obstacle to AI adoption may not be the technology itself, but rather the task of convincing businesses that their most valuable data is safe enough to use.